Re: nftables: How to stop further chain traversal after accept verdict

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



What I am supposed to do in the chain where i can jump? How can I make
it final verdict - same problem as current chain.
Jump will come back and continue next line.
Goto will not.

>> that's the whole point of chains at all
This means unless rule is not in final high priority table chain we
cannot expect it to be accepted. This also means I cannot have
multiple tables for each team/ project?

On Wed, Jun 7, 2023 at 9:19 PM Tushar Shinde <mtk.tushar@xxxxxxxxx> wrote:
>
> What I am supposed to do in the chain where i can jump? How can I make it final verdict - same problem as current chain.
> Jump will come back and continue next line.
> Goto will not.
>
> >> that's the whole point of chains at all
> This means unless rule is not in final high priority table chain we cannot expect it to be accepted. This also means I cannot have multiple tables for each team/ project?
>
> On Wed, 7 Jun 2023 at 6:27 PM, Reindl Harald <h.reindl@xxxxxxxxxxxxx> wrote:
>>
>>
>>
>> Am 07.06.23 um 13:15 schrieb Tushar Shinde:
>> > What is use of table chains if they are not decision makers in accept case?
>> > We mostly know what to accept and deny everything else. Is this cannot
>> > be done with nftables?
>>
>> exactly the same as in iptables for example PREROUTING it don't matter
>> if you have a ACCEPT rule or nothing matches which would DROP a packet -
>> default policy is ACCEPT
>>
>> but just because it passed PREROUTING/MANGLE/RAW by ACCEPT it there
>> don't mean you want to accept the package in FORWARD/INPUT chains
>>
>> that's the whole point of chains at all
>>
>> Chain PREROUTING (policy ACCEPT 64M packets, 4408M bytes)
>> num   pkts bytes target     prot opt in     out     source
>> destination
>> 1     799M  827G ACCEPT     all  --  *      *       0.0.0.0/0
>> 0.0.0.0/0            ctstate RELATED,ESTABLISHED
>> 2      44M 2500M INBOUND    all  --  wan    *       0.0.0.0/0
>> 0.0.0.0/0            ! match-set EXCLUDES_IPV4 src
>>
>> > On Tue, Jun 6, 2023 at 9:29 PM <seentr@xxxxxxxxxxxx> wrote:
>> >>
>> >> You can try to set mark as well as accept them. This way you will be able to determine in the following chains that some of previous ones had accepted it.
>> >>
>> >> 2023-06-06T15:40:03Z Tushar Shinde <mtk.tushar@xxxxxxxxx>:
>> >>
>> >>> Hi,
>> >>>
>> >>> I have asked the same question to super users. But I looks like I need
>> >>> help from experts.
>> >>>
>> >>> https://superuser.com/questions/1787416/nftables-how-to-stop-further-chain-traversal-after-accept-verdict
>> >>>
>> >>> Question is, If there are multiple tables (possibly owned by multiple
>> >>> teams/projects ), if a rule is accepted by one of the tables in the
>> >>> chain, how to stop looking into the next high priority chain. Right
>> >>> now my observation is ACCEPT just stops processing the current chain
>> >>> and jump to next table. If this is last table then only ACCEPT will
>> >>> work, otherwise this ACCEPT is useless.
>> >>> Logs are on SU post.
>> >>>
>> >>> Kindly correct me if I am wrong.




[Index of Archives]     [Linux Netfilter Development]     [Linux Kernel Networking Development]     [Netem]     [Berkeley Packet Filter]     [Linux Kernel Development]     [Advanced Routing & Traffice Control]     [Bugtraq]

  Powered by Linux